It's a really very nice idea to filter strings with string labels, so I suggest to add more details about the context (such as, audio, graphics, earmuffs, mic, avatar culling, etc) into specific string labels, which provide more categories and useful info. To do these, I have some work logs created by Simplified Chinese proofreaders, these screenshots shows some of them. These should be much more friendly to other localization languages.
Full data pack were shared with official devs.